Author Janine Dahinden is professor of Transnational Studies at the University of Neuchâtel. Her research interests cover topics regarding mobility, migration, ethnicity, transnationalism, social networks, gender and culture. Her research is transdisciplinary and combines qualitative and quantitative methods (see also www.unine.ch/janine.dahinden). [email protected] © 2010 by the author ISSN : 1662-744X

Abstract In this article, I am interested in the different types of boundaries emerging in a city that is characterized by a highly diverse population. The analysis of the personal social networks of 250 inhabitants of a small Swiss city – different types of migrants as well as non-migrants – supplemented by data from qualitative interviews, brings to light the relevant factors for the network boundaries, the social positioning, and cognitive classifications of its inhabitants. The population is organized and structured along certain dimensions that reflect patterns of boundaries as an interplay of nationality, place of birth, education, local establishment, mobility type, religion, “race,” and transnationality. It becomes clear that the common ideas of assimilation cannot grasp the complexity of the “categorical game” at place in this city when it comes to migrant’s (and nonmigrant’s) incorporation.

Keywords : Cities, ethnicity, boundary work, social capital, Switzerland, transnationality

Acknowledgment Funding for this research was provided by the Swiss National Science Foundation. I thank René Schaffert for statistical support; Ellen Hertz, Christin Achermann, Kerstin Dümmler, and Marylène Lieber for their important and helpful comments, which have certainly enhanced this manuscript. A first version of this paper was discussed at the AHRD Diasporas, Migration and Identities Programme/CRONEM conference 2009, Diasporas, migration and Identities: Crossing Boundaries, New Direction, University of Surrey, Guildford, UK, 11-12 June 2009. I also thank Susanne Wessendorf who invited me to present a revised version of paper at the workshop entitled, “Researching “Super-diversity” in European cities: conceptual and methodological challenges” at the IMISCOE/SUSDIV annual meeting held September 8-11, 2009 in Stockholm, Sweden.

3

1. INTRODUCTION Cities are, by definition, places of intensified diversity and heterogeneity. People with different lifestyles and socioeconomic resources, diverse occupational, linguistic, religious, ethnic, or national background meet, socialize, or maybe segregate. One of the main features of cities is that they are – and always have been – both locations and outcomes of immigration and integration. It is not by coincidence that migration sociology had its birth at the beginning of the last century in the rapidly growing city of Chicago and that urban anthropology was founded by the last midcentury after anthropologists were obliged to follow “their” rural migrants into the growing cities of Africa. Since then, different generations of social scientists have been engaged in theorizing the issue of how cities are – or should best be – organized in terms of migrant incorporation. Until this very day, academics (and politicians) conceptualize cities, in general, as composed of a mosaic of ethnic or nationally (and in the US “racially”) defined groups, each with its inherent characteristics and its own dense fabric of ethnic organization and clearly demarcated boundaries. The question is then either how the ethnically defined groups assimilate into the main society (assimilation paradigm) or how the cultural specificities of such groups could be recognized and valorized allowing an immigrant’s full participation in national societies as cultural minorities (multicultural incorporation paradigm). Those common narratives have recently been challenged by different theoretical arguments and empirical case studies. For our purpose, I would like to mention three: First, diversity in cities is nowadays enhanced by the restructuring processes of globalized economies and by more intensified forms of embeddedness of cities in transnational spaces (Dahinden 2009, Glick Schiller and Caglar 2009). Vertovec (2007) coined the term “superdiversity,” meaning a condition characterizing cities today and distinguished by a dynamic interplay of variables among an increasing number of multiple-origin, transnationally connected, socioeconomically differentiated and legally stratified immigrants. Hence, the issue of the “integration of difference in cities” got under those circumstances a new actuality. Second, scholars pointed to the ways migrants continually negotiate and renegotiate the terms of their presence, places, and citizenship in cities, demonstrating that societies are more variable, migrants more proactive, and cities more meaningful than current public debates swirling around such concepts as multiculturalism and assimilation might suggest (Hanley et al. 2008). And finally, such “community studies” came in a while under fire not only for their tendency to “groupism” (Brubaker 2004) by treating ethnic groups (as well as nations, and “racially” defined groups) as substantial entities to which interest and agency can be attributed, but also for their underlying methodological nationalism (Beck 2002; Wimmer and Schiller 2002). Such objections underscore the constructed, relational, subjective, and interactional nature of ethnicity and refer to the works of scholars like Frederik Barth (1969) or Max Weber (1996 [1922]). Instead of essentializing “ethnic” immigrant groups by taking them for quasi natural starting point for all investigations, academics propose in this line of rationale to focus on ethnic boundary work (Alba 2005; Dahinden 2008; Wimmer 2009) or on processes of ethnic selfidentification and external categorization (Jenkins 1997) to understand the incorporation of migrants and non-migrants. In this vein of reasoning, this article is interested in the types of boundaries emerging in a city characterized by high immigration. The aim is to demonstrate, through the prism of the social networks of the inhabitants – migrants and non-migrants – of Neuchâtel, a small Swiss city, which categories are brought up in such boundary work and the place ethnicity has among them. In other words, the theoretical and empirical ambition of the paper is to give insights into the benefits we may have when analyzing migrant incorporation processes with such a boundary and network perspective.

Social network analysis, which has a long tradition in urban anthropology (Mitchell 1969), is a framework that suits this research question about the nature of boundary work very well: the focus is placed on the structure of social relations rather than on preliminary (ethnically) defined groups, and this encourages the exploration of multilevel and crosscutting ties and allows “unbounding” problematic concepts like “ethnic or ethno-national groups”. I investigated 250 personal social networks of inhabitants of Neuchâtel to analyze their structures in terms of different types of boundaries. In a second step, I conducted in-depth qualitative interviews that complete the network analysis and enhance our understanding of classifications of the inhabitants relevant for the network boundaries. The idea of boundaries has come to play a key role in important new lines of scholarship across the social sciences. For the actual purpose, I distinguish between network boundaries, symbolic and social boundaries. Network boundaries are structures of membership and therefore exclusion and inclusion emerging out of the personal networks of the inhabitants. Following Lamont and Molnar (2002:168), I understand by symbolic boundaries conceptual distinctions made by social actors to categorize objects, people, and practices. And still following Lamont and Molnar, social boundaries are defined as objectified forms of social differences manifested in unequal access to and unequal distribution of resources and social opportunities. In this line of thinking, I argue that the inhabitant’s network structures display specific network boundaries that are translated into demarcated symbolic boundaries - sometimes leading to social boundaries. By this modality, the population of Neuchâtel is organized into stratified social positioning, where an interplay of nationality, education, local and transnational anchorage, “race,” and religion are the most important structuring forces. In the first section, a short description of the city of Neuchâtel is provided before some relevant concepts of network theory are introduced. Then, the methodology is briefly outlined after which the network structures of the inhabitants of Neuchâtel are presented. Afterward, it is demonstrated based on the results of the qualitative interviews how network boundaries are declined into symbolic boundaries. The final chapter discusses the lessons we might learn for the analysis of migrant incorporation in cities when going beyond ideas of assimilation applying such a boundary making perspective and a network approach.

2. NEUCHÂTEL: A CITY OF IMMIGRATION It might be on a first view astonishing that a small city like Neuchâtel of roughly 32,000 inhabitants, located in the French-speaking part of Switzerland, should be of interest for a larger public1. However, a short glance at its immigration history and at the composition of its population shows that this small city might be worthy of academic attention when it comes to discussing urban life, transnationalism, and migration dynamics. From the 17th century onward, Neuchâtel was touched by an economic globalization that incorporated the city and its surroundings in an increasingly transnational space: growing industrialization (bobbin lace, calico printing), later on the boom of the (famous) watch-making industry, and finally the far-reaching economic restructuring since the 1980s, with new industries (like luxury watches, micro- and biotechnology, medical technology) are the three most important developments (Berset and Crevoisier 2006). These developments sparked the immigration of workers – skilled and non-skilled – from other corners of the world, and Neuchâtel came to accommodate different types of migrants. First, at the beginning of the 20th century and after World War II, it received French, German, Italian, Spanish, and Portuguese migrants, but also migrants from neighboring cantons, seeking employment; then, in the latter half of the 20th

3

century, Neuchâtel also became home to dissidents of the communist regimes of Eastern Europe, refugees and asylum seekers from former Yugoslavia, the Middle East, Asia, and Africa. Given the multinational character of the new industries built up since the 1980s, they attracted highly qualified workers with skills in management and in specialized tasks, and a transnational business network was established. However, since 1989, Switzerland has replaced its relatively liberal admission policy by a more restrictive one. EU citizens and members of the EFTA countries are increasingly freely admitted; for non-EU citizens, on the other hand, the new policy offers educational and work opportunities almost exclusively to executives and other highly qualified individuals. At the same time, and in line with most other European countries, Switzerland is construing its asylum policy in increasingly restrictive ways (Piguet 2005). This history of globalization and the general tendencies in Swiss immigration policy are mirrored in the profile of the population of Neuchâtel: The population is highly heterogeneous with regard to the spectrum of national origins, but also with regard to the length of stay, the different types of residence permits, and the educational background of these migrants (see table 1). Migration has also a strong effect on the composition of the population in religious terms. Historically, the population of Neuchâtel was Protestant – incorporated in the old aristocracy of Neuchâtel, that is, local families or Huguenots who had been ennobled by the French princes or by the Prussian King in the eighteenth century; these people held political and economic power and actually governed Neuchâtel until the revolution in 1848. As a result of the migration flows during the last decades, Catholics – mainly immigrants from Southern Europe as well as from other Catholic Swiss cantons – now represent 39% of the population and outnumber the Protestants who make up only 31%. In addition, 3% of the inhabitants are Muslims. This way, we can say that Neuchâtel is an immigrant city, similar to big cities. However, the small size of this city is, as a consequence, organized differently from other urban areas. For instance, at first glance neither the structure of neighborhoods nor the housing markets reveal tendencies toward separation along national or ethnic lines in Neuchâtel. Unlike big cities such as London or Paris, there are no suburbs populated almost exclusively by immigrants, nor do we find ethnic spatialization with regard to housing. There are some streets within a concentration of immigrant populations, but overall, diversity is organized in a different manner in this small, but still urban, place and has to be structured within a densely inhabited territory. To give an example: there are streets where we find expensive apartments with a beautiful view of the lake and the Alps on one side, while the other side of the street is inhabited by people from popular milieu or of immigrant origin. This is because on the “other side of the street” the houses look out not onto the lake, but only onto the backs of the houses that have the view. Thus, we may ask how the population is structured in terms of social networks, network boundaries, and social classifications.

3. SOME NETWORK THEORY ASSUMPTIONS Network analysis is especially suited to our research question as it allows for theorizing the issue of boundary work while at the same time it can empirically grasp it. Therefore, some of the principal ideas of network research and theory with regard to our topic of interest, hence the creation and organization of different types of boundaries within the population, shall shortly be introduced: The basic premise of network analysts is that the social embeddedness of actors in a web of specific relationships says something about their position in society. In contrast to current approaches network researchers do not regard social systems as a collection of isolated actors with certain characteristics. Rather, their attention is directed toward examining the relations of the actors in a

4

social network and the attempts to describe this pattern. In this way, one hopes to gain information about the possibilities and constraints that affect the actors’ scope of action (Schweizer 1996). These patterns of embeddedness in social relations do not emerge by chance, but should be regarded as structural patterns and are therefore intrinsically linked with the possibilities, as well as the constraints of social action of actors; thus, they influence the resources available to these actors (Scott 1991). With regard to boundary work, two main themes are salient: First, we need to investigate the way in which embeddedness in social relations does implicate specific resources for the actors. Here, it is the guiding principle of network range and diversity as a form of network social capital that is decisive. Are they different groups of people possessing different forms and volumes of network capital? How can the boundaries between these groups be described? Second, we have to elaborate on the way in which embeddedness in social relations produces network boundaries along the so-called homophilic principle. I would like to develop the two guiding ideas more in detail: The first set of theoretical orientations focuses on network diversity and range and therefore on a specific asset of social capital. In the past decades, social capital in its various forms and contexts has emerged as one of the most salient concepts in the social sciences. Lin (1999), and later Flap and Völker (2004), have outlined a network theory of social capital based mainly on the works of Bourdieu (1980) and Coleman (1988). Network social capital here basically refers to the importance of resources, which, although possessed by others, are available to a given individual through his or her social relations with those who own them. Lin (1999:35) defines social capital “as resources embedded in a social structure which are accessed and/or mobilized in purposive actions”. Following Lin, in this article, network capital is considered as one possible form of social capital and asset in a network.2 The core idea of network capital is that people better equipped with social resources – in the sense of their social network and the resources of others that they can draw upon – will succeed better in attaining their goals. There is general agreement in the literature that this kind of network capital varies with network size and network composition (or ‘volume’ and ‘structure’) (Burt 1983; Van der Gaag and Snijders 2004; Degenne and Forse 1994). The importance of size is obvious. Ego’s network has range, to the extent that it connects him/her with a diversity of other actors. However, measuring range only as the number of ego’s contacts requires the assumption that each of ego’s contacts equally increases the range of his/her contacts – an assumption that is – following Burt (1983) – obviously problematic. At this point, the composition and quality of contact comes into play: A contact has quality to the extent that it increases the diversity of alters in ego’s network. For instance, with regard to fractions of “weak” versus “strong” ties, the underlying idea is: While weak and bridging ties are valuable instrumental resources (access to information, to jobs etc.), strong and emotional ties are closely associated with social support and solidarity, representing two different aspects of network social capital (Burt 1982 ; Granovetter 1973). In other terms, the pivotal idea is that the more diverse a network is in terms of its composition (or structure), the more social capital is embedded in the same network. This is where calculations of heterogeneity indices over the network come into play: Networks with high diversity that include diverse ties (“strong and weak ties”) and a broad range of ties with qualitatively different connections to diverse others (in terms of gender, ethnicity, class, or more general, in terms of status) but also in terms of roles (kin, friends, etc.) do embed more network social capital than networks that are characterized by low diversity and range. Meanwhile, it is common to use name-generators to measure social capital (for instance, Van der Gaag and Snijders 2004, Moore 1990), a procedure that has been adapted in this study as well (see chapter methodology).

5

Going back to the question of the network boundaries characterizing the population of the city in question, we may ask if they are individual inhabitants who display similarities with regard to the social capital embedded in their networks, and if they are specific persons who have similar or different amounts of network social capital and if this is the case, by which characteristics they could be described. The second vital question is the following: Do the networks of the inhabitants follow homophilic principles and therefore display network boundaries along certain variables? We speak of homophily when a more-than-coincidental similarity in characteristics and attitudes can be observed in people who are linked together by a certain kind of social relationship. Aristotle complained that people “love those who are like themselves,” and this popular adage has been confirmed empirically by many studies (Lazarsfeld and Merton 1954): Homophily along ethnicity created the strongest divides in western personal environments, with age, religion, education, occupation, and gender following in roughly that order (McPherson et al. 2001: 416). Homophily has far-reaching consequences for social action and cognition: The most important aspect is that homophily limits people’s worlds and has powerful implications for the information they receive, the attitudes they form, and for their everyday interactions. To put it differently, homophily implies that distance in terms of social characteristics translates into network distance and vice versa. This way, the question we have to answer is: Who sticks (in terms of networks) together with whom and according to which criteria do different segments of the inhabitants stick together? For this study, the principal issue is if it is ethnicity or other variables (gender, age, etc.) that are more important for the grouping of the population. In sum, two kinds of network boundaries can be tagged by these concepts: First, emerging boundaries between individuals possessing of different volumes and structures of network social capital and second, boundaries in homophilic terms of “groups sticking together”.

4. METHODOLOGY To understand the organization of the boundary work at stake in the city, the research design involves two stages. First, I empirically grasped the everyday networks of 250 inhabitants of Neuchâtel using a multiple name generator consisting of 10 items, represented by 10 different questions.3 Using this name generator, the 250 persons interviewed mentioned a total of 3,014 reference persons. Using “name interpreters,” background information was collected about these reference persons. I asked, for instance, about sex, nationality, place of residence, profession, educational level, country of origin, and so on of the mentioned person. Assessing the quality of the relationship between the interview partners and the persons mentioned was also of interest: thus, the respondents were asked to indicate how they were connected to the persons named as well as the length of relation and the intensity and frequency of contact. The sample was drawn from a list of names (people between 20 and 60 years old) delivered by the Residents’ Registration Office in Neuchâtel. The Office keeps a record of every inhabitant living in Neuchâtel, with the exception of asylum seekers, protected persons without long-stay permits, diplomats, and obviously undocumented migrants. When sub-populations vary considerably as it is the case here, it is advantageous to sample each subpopulation independently. Strata were defined by the criteria of nationality: Swiss, EU/EFTA members, and third-country nationals. Then, random sampling was applied within each stratum. With regard to foreign citizens, I included only first-generation migrants (those not born in Switzerland) to avoid too diversified a sample. The

6

migrant population is over-represented in the sample (see table 1), which is due to the selection process described and which was done intentionally to obtain valuable results with regard to fragments of the population that are less present in the city (such as non-European migrants). For analysis, the data have been statistically weighted, while the overall population of the city (see table 1) served as approximation. Such procedure allows for generalization. Table 1: Sociodemographic characteristics of the overall population (2007) and profile of the sample (network study) Overall population NE in 2007 32’389 (100%) 48% 52% 69% 20% 11%

Men Women Swiss Nationality EU/EFTA Third-Country Nationals 69% Residence Status Swiss Annual permit 9% Residence permit 20% Other 2% Protestant 31% Religion Catholic 38% Orthodox No data Muslim 3% None 20% Other 8% Switzerland Place of birth Outside Switzerland Neuchâtel Outside Neuchâtel Source Data 2007: Statistical office of the Canton Neuchâtel Sex

Profile sample network study 250 100% 129 52% 121 48% 106 42% 63 25% 81 33% 106 57 81 6 47 72 13 23 69 26 98 152 27 218

42% 23% 33% 2% 19% 29% 5% 9% 28% 10% 39% 61% 11% 89%

Table 1 displays the main features of the people surveyed. First of all, the 250 interview partners were diverse as the inhabitants of Neuchâtel in general. With regard to national category, 42% were Swiss citizens, 23% came from EU or EFTA countries, and 33% were citizens of countries outside Europe. 4 The sample was made up a total of 45 different nationalities. Furthermore, it was also heterogeneous with regard to the types of migration (asylum, labor market), education, religious conviction, and length of stay in Neuchâtel and Switzerland. Second, the objective of the qualitative in-depth interviews, which have been conducted in a second step, was to investigate the social classifications, how the inhabitants’ identify with their city and in general to grasp their social identifications be it locally or transnationally. The interviews partners were, for instance, asked to identify the groups that are most important in their eyes – to draw a kind “sociogram”– and to explain the boundaries between these groups, be it in Neuchâtel, in Europe or globally. In total, 18 people were interviewed during this second stage. Achieving variation and saturation were the main concepts concerning the selection of the informants, though following theoretical sampling (Glaser and Strauss 1967). We spoke to Swiss families, working immigrants, naturalized immigrants, and recognized refugees, as well as highly qualified immigrants.

7

The network data were coded according to the characteristics of the interviewees and their reference persons, as well as to the relation between them, and analyzed with different procedures in SPSS. The data from the qualitative interviews were analyzed according to a content-reduction strategy introduced by grounded theory (Charmaz 2001). For analysis, data have been triangulated. Overall, we can state that the network data were used to inductively uncover the salient boundaries while the data from qualitative interviews were used in order to better understand those network boundaries.

5. EMERGING NETWORK BOUNDARIES 5. 1. Network diversity and range In this first part of the analysis, I investigate whether there are individuals or specific clusters of people characterized by high network range and diversity – and though high network social capital – and, at the other end of the scale, others who stand out because of low degrees of network diversity and low network social capital. Network diversity and range was measured using three variables commonly used in network research (Flap et al. 2005; Van der Gaag and Snijders 2004): a) The relative proportion of kin: the higher the proportion of kin in the network, the higher the potential for solidarity and social support, but the less diverse the network – according to a commonly used network hypothesis; The more the diversified the ties in terms of roles in one’s network, the higher the network capital. Respondents were asked to indicate the ways they were connected to each of the persons named. The linkages were then classified into six categories: kin, friends/acquaintances, persons related to work/job, persons from an institution or a professional association, and others. The assumption is that networks consisting of all six categories of ties incorporate more network capital than the ones consisting only of one type. b) Variation in strength of ties: This indicator measures strength and weakness via frequency of contacts (for a discussion about the use of this proxy see Marsden 1990). Respondents were asked to indicate the intensity of contact with each named person using 6 categories: every day, a few times a week, every week, every month, every three months, or less frequently. Thus, diversity is expressed by high values, meaning that in the network all categories (and though strong as well as weak ties) are represented, which enhances the network social capital. c) The overall network size: this means the total number of persons named. The bigger the network, the higher is the chance of diversity and of having access to different parts of society. The total number of possible reference persons was limited for every question in the name generator to 3, the maximum value for network size being 30.5 In the first step, I calculated differences in the three network diversity variables (relative proportion of kin / variation in strength of ties / network size) according to the different items (nationality, place of birth, gender, etc.). Data were weighted for these calculations using the SPSS Complex Samples procedure (simple contrasts for estimated means in the general linear model). In table 2, the values for the three variables according to different items are presented. The analysis shows that some means vary considerably between different items while not being statistically significant. This is why only those items showing statistically significant values are discussed.6 Which items go with high network social capital according to this perspective? Starting with place of birth, it reveals that people born in Switzerland have significantly lower proportions of kin in their

8

networks and higher variation in strength of ties than people born in Southern Europe or outside Europe. On the other hand, persons who had built their lives in Switzerland or in central or northern European countries do not differ between them with regard to these variables. Concerning network size, the values of the Swiss contrast from those born in Southern Europe. In summary, people having their place of birth in Switzerland or in central or northern Europe have significantly higher volumes of network social capital than those born in southern Europe or outside Europe. Table 2: Network diversity and range: Contrasts and variations (data weighted) Relative proportion of kin within the networks

Variation in Strength of Ties (max. 6)

Mean % 38.9

Mean % 4.6

All the whole network

Network Size Mean % 12.4

PLACE OF BIRTH Mean % Switzerland (S)

36.7

Southern Europe (SE)

50.0

Central -Nord Europe (C/N-E)

32.9

Outside Europe

43.5

S

SE

C/NE

**

Mean %

S

SE

4.8

**

** **

4.2

C/NE

**

4.1

S

SE

**

C/NE

* 11.4

**

*

12.2

4.6

*

Mean % 12.7

12.2

**

MOBILITY PATTERN Lived only in Switzerland (CH)

Mean %

CH

O + CH *

37.2

Mean %

CH

O + CH

4.7

**

Mean %

CH

O + CH

WM

AS

12.3

*

in the country of origin and in Switzerland (0+CH)

44.1

Lived in more countries

36.8

4.1 *

**

11.9

4.7

**

13.1

TYPE OF MIGRATION

Working migrants (WM)

Mean % 45.8

Asylum (A)

53.0

No immigration

36.7

WM

AS

**

**

L

M

Mean % 4.0

Mean % 11.8

4.0

14.3

4.8

WM

AS

* *

12.7

**

EDUCATIONAL DEGREE Mean %

Mean %

Low (L)

41.2

4.1

Medium (M)

44.8

4.4

High (H)

34.4

**

4.8

L

M

Mean %

L

M

**

*

10.9 11.7 **

*

13.2

RELIGIOUS CONVINCTIONS

Catholic (C)

Mean % 46.3

C

M

P

Mean %

P

Mean %

**

4.3

**

12.0

3.9

**

12.7

Muslim (M)

44.7

Protestant (P)

33.0

**

5.0

No religious conviction

37.5

*

4.6

C

**

M

13.6

** *

*

A

R

S

*

**

C

M

P *

*

11.8

RESIDENCE PERMIT

Annual

Mean % 42.6

A

R

S

Mean % 4.3

Residence

42.4

4.2

Swiss

37.4

4.7

Mean % 11.5

A

R

S

11.8 12.7

9

INCOME Mean %

CH

O + CH

Mean %

CH

O + CH

Mean %

0 – 3’500 CHF

37.9

4.4

11.6

3’501-5’500 CHF

41.3

4.7

12.8

5’501-20’000 CHF

37.8

4.7

13.4

0-3’500

3’5005’500

*

GENDER

Women Men * significant p < 0.05 ** significant p < 0.001

Mean % 37.2 40.5

w

Mean %

w

4.7

Mean % 12.4

4.4

12.4

W

Furthermore, mobility pattern influence network social capital in a complex way: Data show that migrants having lived in different countries have significantly more diverse personal networks than those having lived only in Switzerland and their country of origin, while no significant variations and contrasts are eligible with regard to network size when it comes to mobility. In the same vein, the network composition of people having spent their lives exclusively in Switzerland displays more volume of social capital than the one for those migrants having lived only in Switzerland and in their country of origin. In sum, we can read that mobility influences network social capital in two ways: Immobility on the one hand and enhanced circulation on the other hand involve both high volumes of network social capital, while movements of one-way migrations implicate low level of network diversity. Similarly, type of migration is relevant: People having arrived in Switzerland as working migrants show higher proportions of kin and lower variation in strength of ties than persons without immigration experience, the latter possessing this way higher network capital than the first. Also, persons having arrived through asylum have less network social capital than those without immigration experience. Education is possibly one of the factors with the strongest influence on network diversification. This rule, confirmed in numerous network studies (for instance, Meyer 2001), can be applied to this study as well: People with high education show higher values for network range and diversity than people with low (though not significant for proportion of kin) and medium education. This result is interesting as it contains that that cultural capital translates itself into network social capital (or vice versa). More astonishingly might be that different religious convictions vary in network social capital: Catholics’ network social capital is significantly low for all three variables when contrasted with the Protestants. Also, with regard to variation in strength of ties, Muslims display significantly lower values than Protestants. The picture emerging with regard to residence permit and income is more blurred: Persons holding either an annual permit or a residence permit show contrasting values with the ones of Swiss citizens concerning tie variation. Income structures personal networks only in one way: Those earning most money differ from those with the lowest salaries with regard to network size. And finally, the network structures do not significantly differ between women and men with regard to social capital. Who are ultimately the persons that display high volumes of network social capital? Being born in Switzerland or in central or northern Europe, possessing high educational capital, being Protestant, immobile or on the other hand highly mobile, and having a high salary are items that are associated with high values for network diversity and range. On the other hand, the following characteristics point to low volumes of network social capital: people from Southern Europe or outside Europe, working migrants or having arrived through asylum, Catholics (and less Muslims),

10

persons having experienced a one-way migration and lived in the country of origin or in Switzerland, and those possessing a residence or an annual permit. In other terms, I argue that we witness first – while still blurred – shadows of network boundaries. These network boundaries are without doubt related to inequality as they embody different volumes of network social capital and give therefore not equal access to subgroups and resources.

5.2. Homophily To affine the shape of arising network boundaries, the analysis of homophily is fruitful. As an expression of the strength of homophily, a correlation coefficient was used for the evaluation of the association between the corresponding characteristics of the interviewees and the mentioned associates. The whole network is segregated, in this order, according to nationality, ethnoregional origin, education, residence status, age, length of stay in Switzerland, and gender (table 3). Table 3: Homophily by Subgroups Valuesa Nationality Ethnoregional origin (Swiss, Southern Europe, Central and Northern Europe, outside Europe) Education

Cramer’s V

ALL .554

Men .617

Women .554

Cramer’s V

.494

.518

.474

Spearman

.432

.384

.485

Residence status

Cramer’s V

.330

.318

.363

Age

Spearman

.286

.265

.305

Length of stay

Cramer’s V

.239

.231

.286

Sex

Spearman

.210

-

-

The possible value of the correlation coefficient ranges from -1 to +1, whereby the sign shows the direction of the association. If we are dealing with dichotomous or ordinally scaled variables, the correlation according to Spearman is used as the association; with nominal variables, Cramer’s V is used.

Stating that the main characteristic of the everyday networks of people living in Neuchâtel is their marked national and ethnoregional homogeneity (tables 3 and 4) I mean, for example, that a large portion of Swiss people turn exclusively to other Swiss in daily interactions, with over 80% of the reference persons named by Swiss citizens in answer to the “name generator question” being other Swiss people (meaning that 20% are non-Swiss). We even can say that the Swiss display the highest value with regard to this item. But also, Europeans from the South as well as thirdcountry nationals interact in over half of the cases within their national categories (57% and 52%, respectively). Central and northern Europeans have a proportion of 41% of the persons of the networks coming also from this region. Interestingly enough, the data reveal also that, for instance, people from Southern Europe do not stick together with the ones from Central and Northern Europe or from outside Europe: Rather, in case the mentioned associates do not belong to their own ethnoregional origin, they are mainly Swiss. This result is without doubt affected by baseline homophily, due to the demography of the potential tie pool (there are more Swiss than others in Neuchâtel, around two-thirds of the resident population, while there are for instance, only 11% third-country nationals (see table 1)). Nevertheless, we can say that network boundaries along nationality and ethnoregional origin are observed in the networks of Neuchâtel inhabitants.

11

Table 4: Homophily by ethno- regional category National categories Central and Southern Northern Europe Europe

Swiss Swiss

Count %

Southern Europe

Count %

Central and northern Europe

Count

Outside of Europe

Count

% %

Total

Count %

Outside Europe

Total

795

49

75

54

973

81.7%

5.0%

7.7%

5.5%

100.0%

151

236

15

11

413

36.6%

57.1%

3.6%

2.7%

100.0%

215

20

178

25

438

49.1%

4.6%

40.6%

5.7%

100.0%

321

40

53

452

866

37.1%

4.6%

6.1%

52.2%

100.0%

1482

345

321

542

2690

55.1%

12.8%

11.9%

20.1%

100.0%

Furthermore, and this is of specific interest, almost as important as the national category in structuring the everyday interaction networks of inhabitants of Neuchâtel is the educational attainment, this value being just slightly lower than that the one for ethnoregional origin (table 3): better educated people rely on other persons with a good educational level and stick together with them. Sixty-eight percent of the networks of people with high education are composed of associates possessing also high education; among the lowly skilled interview partners 40% of their reference persons are also lowly skilled (not shown in the table). Concerning the migrants, the type of residence permit reveals to be another important structuring criterion. In Switzerland, as in many other countries, different types of legal permits imply different sets of social, economic, and political rights or, conversely, exclusion from such rights. If migrants remain in their daily interactions, for whatever reason, within the boundaries of their permit, the permit type can become a means of exclusion as the resources of the associates in the networks will be limited. Finally, we again find the result that the tendency for homophily with regard to gender is not very marked, at least when compared to the other criteria. To summarize, the data show that the heterogeneity of the population of this city is translated into different patterns of specific groups sticking together: These groups are grosso modo built upon the same items we identified for network social capital, producing therefore a kind of “double” network boundaries. Educational level and nationality are not only articulated with different amounts of social capital embedded in the networks as shown in the previous paragraph, but they go also along with cohesive forces among those who share the same characteristics within these categories, be it to be born at the same place or to being highly or lowly skilled. In other terms, having more or less network social capital results not only in groups with demarcated network boundaries, but those boundaries are also reinforced through homophily tendencies: Those with high network social capital stay among themselves and they can profit from the multiplier effect of social capital – those having modest network social capital stay with other with the same characteristics, turning mute this multiplier effect of social capital. Put differently, the different forms of homophily displayed by the data mirror the practice that groups of people turn to

12

persons with similar amounts of personal resources in terms of education or residence permit as well as network social capital. We can therefore assume that persons with limited personal resources and capital – for instance, low education and therefore low human capital and / or residence permits with limited rights and low network social capital – turn to other people with similarly limited personal capital. This could mean that those persons have only limited possibilities for accessing the society’s resources, at least through their social network. To conclude, I would argue that these results point to different hierarchical social positioning of these groups in Neuchâtel.

5. 3. Emerging types: clusters of people mirroring those network boundaries In the previous paragraphs, different variables have been identified that create the lines along which boundary work takes place in terms of network social capital and homophilic tendencies. Out of these results a typology can be established. This typology has to be understood in line with (Max Weber1991 [1904]) as a means to grasp and understand conceptually social phenomena. A type is here a model of an abstract nature and serves the purpose of theory building. The types were built up according to the general pattern made evident in the previous two paragraphs. The aim was to identify specific clusters of persons (with high mobility, highly educated, born in Switzerland, and so on) who show different network boundaries. Theoretical considerations also guided this “inductive” search for patterns. On this basis, we can identify tentatively four different types of clusters of people reflecting different network boundaries and incorporating different volumes and compositions of social capital. These “groups” are differently positioned in the city and stratify the urban space. The first type is composed of people born in Switzerland with no migration experience. We are speaking here of the immobile ones having high network social capital. Those entering into this group have mainly lived in Switzerland and are Protestants. Persons with such characteristics show a high degree of diversification in their networks, and they “stick together” according to the analysis of homophily. Out of this description, I would like to formulate the hypothesis that it is local and historical anchoring, which is the main driving force of both the specific network composition and the appearing network boundary. I will label therefore this first type “the local old established.” Another type that displays high network social capital may include persons born in central or northern Europe who lived in different countries, have high educational background, and earn considerable salaries. The mobility of these persons cannot be traced by a unilinear movement from the country of origin toward Switzerland: on the contrary, they have lived in different countries before coming to Switzerland. It is their mobility and their high cultural capital that distinguishes this type from the others. I will call them the “the highly skilled mobiles.” Interestingly enough, the values for the network diversity are in this case similar to those of the “local established”; however, the structure of networks of these ideal types might differ with regard to several fundamental modalities. Whereas the “local established” constitute their social capital through “locality” – being born and raised in Neuchâtel and having no migration experience, the “highly skilled mobiles” constitute theirs through high education and mobility. The third type could be described as follows: Immigrants, born in Southern Europe, Catholics, working migrants who have a residence permit, characterized by low education, persons who lived either in their country of origin or in Switzerland but at no other place, display network with low diversity values. Being familiar with Swiss immigration history, it reveals that we are dealing here with characteristics of the traditional, so-called guest workers who arrived after the Second World War and have settled down in Switzerland while they are still transnationally connected. I labeled this group “the established guest workers.” Mobility is also salient, as was the case with the second

13

type, but it takes on a different shape. What distinguishes them from the “highly skilled mobiles” – besides their more homogeneous networks – is their settled character and their lower educational background. Finally, the last type could include persons with the following characteristics: they are born outside Europe, are often Muslims, have arrived in Switzerland seeking asylum, and obtained then an annual permit. I call them “the recently arrived outsiders.” Their network structure points to modest social capital and they keep to themselves in quite a cohesive manner: and as the associates in their networks are in a similar situation, the multiplier effect of social capital will in this case be mute. In sum, the first two types point to persons having high levels of social capital embedded in their personal networks, whereas their capital might be differently constructed (local anchorage versus mobility and high educational capital). The last two maintain in their personal networks lower volumes of social capital – and again, the structure of the networks indicates that underlying forces might be different (settled down – newly marginalized). In the next paragraph, these types shall be described more detailed based on the results of the qualitative interviews.

6. SYMBOLIC BOUNDARIES REINFORCING AND (RE)PRODUCING NETWORK BOUNDARIES It is evident that in this small city the network structures of its inhabitants and the emerging boundaries are closely articulated with questions of resources (material and immaterial): By including theoretical reflections and results out of the qualitative interviews, we are able to better understand the formation and boundaries of those clusters of people. It reveals that the salient network boundaries identified in the last chapter are mirrored in social classifications and distinctions – symbolic boundaries: hereby, the degree of establishment and historical anchorage, education, and mobility as well as stigmatization according to “race,” and Islam play an important role. In designating the first type with the notion “established” – “the local established” – I refer to the theory of figuration of Elias and Scotson (1965). They pointed to the importance of old established groups when newcomers arrive and demonstrated how the established groups close ranks and reinforce internal cohesion, to keep the newcomers at the bottom of the social hierarchy and out of their circles. The means of exclusion are the cohesion of the group, as well as stigmatization, humiliation, and gossip. Based on the network structures and the results of the qualitative interviews, I shall maintain that something similar is occurring in this small city: Old established native families not only close ranks against newcomers, but they also seem to be able to profit from historical grids of power relations. The offspring of Protestant aristocratic families who held political but also economic power till the revolution in 1848 and all other families who got established during the last generations seem to be able to profit from anchorage and local advantage and closing their ranks toward newcomers as the data show. Thus, even in a globalized world “locality” – meaning the concrete local and historical constraints and structural opportunities – retains its influence on the organizational patterns of the inhabitants of this city. Their potential for accessing diverse resources and their social capital is grounded in traditional continuity and local advantage. But the “established transnational guest workers” also have this characteristic of being locally anchored – at least when compared to the newer immigrants. They make also use of both demarcation and closure to maintain their positions. In other terms, they are similarly “established”

14

in its double sense of a (short) historical anchorage and as a means of excluding others. The immigrants from Italy, Spain, and Portugal arriving since 1950 have experienced upward mobility; they have accommodated themselves with their families in Neuchâtel and have children of the second or third generation. Like a representative of the guest workers told us: “ W ith regard to their passports, there are many Italians, Spaniards, Portuguese, but they came 40 or 50 years ago and they made themselves Neuchâtelois. They are integrated. If you take the names of the city’s entrepreneurs, they are all Italian names. ” Conversely, I would formulate the hypothesis that these immigrants have been able to establish themselves locally not least by passing on to others the ball of marginalization and discrimination that was theirs in the 1960s and 1970s and by closing their ranks toward the new immigrants. One of the Italians interviewed, who has lived in Neuchâtel for 30 years, said: “ We, the Italians, also experienced terrible things, discrimination, and so on. Now, this kind of thing is happening to the newly arrived ” .

When these immigrants arrived they were at the bottom of Swiss society, but they experienced upward mobility as they were replaced by new groups of immigrants. It becomes clear that they do not mix with the newly arrived not only from the network data (see homophily), but also from the qualitative interviews. Most first-generation Italians and Portuguese interviewed told us that they do not have anything to do with those who arrived more recently – Turks, Africans, and so on. Similar processes of demarcation between new and old immigrants have been reported by other research in Switzerland – which not surprisingly, follow similar lines (Wimmer 2004). Such demarcation is not of a casual character, however, as this “distinction” (Bourdieu 1982) sometimes goes along with stigmatization and external categorization – an element which, according to Elias and Scotson, is highly relevant in processes of boundary closing. All refugees interviewed, representing the “recently arrived outsiders,” mentioned that stigmatization emanates from the old immigrants, rather than from the Swiss. Here is just one voice to illustrate this: “ T he paradox is that racism in Neuchâtel is something that often comes from the old migrants: the Portuguese and the Italians versus the Africans, the Arabs, or the Turks. ” It should be emphasized, however, that stigmatization of the “recently arrived outsiders” emanates not only from the established guest workers, but also from most of the people interviewed – be they Swiss, highly skilled or established immigrants. In response to our question in the qualitative interviews about whom they would prefer their children not to marry, the majority of the interviewees answered that they would not like their children to marry a Muslim or a black person. In other words, they mentioned the characteristics ascribed to the members of this fourth type “racializing” hereby boundaries. Thus, “race” and “Islam” serve as a means of stigmatization, conceptual distinction, and external categorization, closing the network borders and segregating the “recently arrived outsiders” from the rest of the inhabitants of Neuchâtel.

15

What about the last group, the “highly skilled mobiles”? As briefly mentioned in the introduction, the (famous Swiss) watch was 1970 and 1980 hit by a severe crisis due to competition from Japanese and Hong Kong firms that had introduced electronic production methods. The crisis triggered a deep economic restructuring as cantonal as well as federal politicians began to promote economic development by a policy attracting foreign firms by specific measures (tax breaks and financial incentives for foreign companies; residence and work permits for foreigners; help in finding houses, apartments, and schools for their children; and so on). This policy succeeded insofar as industries of multinational stature mushroomed that attracted highly qualified workers for management, specialists in design and marketing, and other technological competencies not available in Switzerland. These people are highly embedded in transnational networks and the interviews show that they are delocalized and are not anchored very well in the city. From the interviews the image of a kind of “satellite” emerges: They do not know the city very well, they are not even in a position to say anything about the composition of Neuchâtel’s population. In the interviews, the people representing this ideal type were often unable to answer the question asking them to describe the city. This is a global elite circulating and integrated in transnational networks and not incorporated into the local structures of the city.

7. GOING BEYOND ASSIMILATION? The network perspective on incorporation processes puts forward the idea that the inhabitants of Neuchâtel are organized and structured along certain dimensions reflecting clear boundaries that result from an interplay of nationality, place of birth, education, local establishment, mobility type, religion, “race,” and transnationality. Network boundaries are reinforced by symbolic boundaries and translated into social boundaries pointing to social hierarchies and social exclusion among the inhabitants of this City. Overall, the boundary works leads to four clusters of people: the ‘local established’, the ‘established guest workers’, the ‘highly skilled and mobiles’ and the ‘new outsiders’. Those patterns are historically grown, complex, transformative and interactive. The questions that need to be addressed in this conclusion are: Is there a way that such complex processes of inclusion and exclusion could also be grasped with the concepts of assimilation? What potential new insights could we gain from such a boundary and network perspective on migrant incorporation processes? In fact, the network idea is not alien to assimilation theory. The degree to which members of immigrant groups forge primary relations with native-born members of other ethnic groups constitutes the linchpin ideal in traditional assimilation theory. Milton Gordon (1964:70) called the development of such primary relations “structural assimilation”, defining it as entering “fully into the societal network of groups and institutions” of the host society. Those ideas have been reexamined over the last two decades and newer conceptualizations recast assimilation less sequentially; it is more segmented, is an outcome of continuous contacts between the members of different groups and is reinforced by institutions and structural factors occurring over generations (Alba and Nee 1997; Portes and Rumbaut 1996). With regard to networks, recent work has shown that communities with closed ethnic networks can provide social support and social capital (Portes 1998; Zhou 1997), counterbalancing the idea of “straight-line assimilation” that being embedded in ethnic networks would always and automatically be a sign of non-assimilation. Nevertheless, I argue that assimilation theory still has important shortcomings. We have to go beyond those ideas when we want to grasp processes of boundary work resulting in insiders and outsiders and when we aim to understand incorporation in local and transnational spaces.

16

The first deficiency of assimilation theory is that it still sticks to what has been called a “containermodel of society”, taking society within the boundaries of a given nation-state as its implicit and natural entity for research (for instance, Levitt and Glick Schiller 2004). This presumption has important corollaries. One is that incorporation into a transnational space – as it is has been brought to light for the ‘highly skilled and mobiles’ – cannot be analytically taken into account. However, the “national lenses” of assimilation theory are not just too narrow to give adequate consideration to transnational aspects. The national prism is, at the same time, paradoxically too broad to grasp “locality”. Local opportunities and constraints – in political, social and cultural but also discursive terms – include always a national dimension, but go clearly beyond it. Local anchorage and local historical factors – for instance, place of birth or religious conviction – are important elements for boundary-making in Neuchâtel. A second pre-assumption of assimilation theory is that it considers one-way migration followed by sedentariness as the ‘courant normal’. However, the results of the study show that different types of mobility and movements have indeed different effects on incorporation and boundary processes – think again of the ‘highly skilled’. Assimilation theory unintentionally ignores an important part of the social realities taking place in a transnational and interconnected global world. Finally, assimilation theory incorporates the idea that “ethnic” or “ethno-national” groups are the natural starting point for investigation and that ethnicity would be the most important criteria for incorporation processes. In other terms, assimilation theory does not analytically distinguish between community, ethnic group and culture, assuming that in the boundaries of an ethnic community, its identity and culture would coincide in an unproblematic way. This is what Andreas Wimmer (2009) recently critically called the Herderian heritage and common sense of migration research. Going back to our study we might ask, what role does ethnicity play in the boundary processes observed in Neuchâtel? It appears that ethnicity and nationality are very relevant for the construction of symbolic and social boundaries and for the network structure: In other words, ethnicity and nationality matter – but they matter in another way than what assimilation theory presumes. They are not the starting point – the essence – for social processes as assumed in assimilation theory, but they are already the outcome of social processes (again going back to the ideas of Max Weber and Frederik Barth) to which a range of different actors – among others, the nation-state – contribute. More importantly, the results show that ethnicity and nationality are relevant to boundary work solely in combination with other categories – for instance, with education or duration of stay. In the end, we have to deal with a kind of ‘categorical game’ which is at stake at this City, and where ethnicity plays a role which cannot be grasped with the common ideas and implicit presumptions of assimilation. An important limitation of this study is that second-generation migrants are not included and that in general, only little can be said about how boundary-making – and unmaking – is taking place in a long-term perspective. Therefore, we have to go back to the field and carefully examine how this categorical game is modified (or not) and find out which elements become more or less salient in due course. When it comes to understanding incorporation in a longitudinal perspective, some ideas of assimilation theory can be very precious, especially its focus on second- (and third-) generation and social mobility.

NOTES 1) In comparison with European or North American cities, the number of inhabitants is very small. However, the reader should keep in mind that Switzerland is a country of roughly 7.5 million inhabitants with only 5 cities of more than 100,000 inhabitants. The district of Neuchâtel has roughly 51,000 inhabitants.

17

2) There are obviously other conceptualizations, measurements, and mechanisms of social capital that are not directly related to the social networks of individuals. I do not, however, discuss them here. 3) The questions were as follows: 1. Most people do discuss important matters with others. With whom did you discuss important matters during the last six months? 2. Is there someone who turns to you if he or she needs to discuss something important? 3. From time to time questions arise with regard to financial issues (for instance, taxes, subsidies and scholarships), insurances (health or life insurance) or specific laws. Whom did you turn to the last time you had this kind of questions? 4. How did you find your current / last working place? Did somebody give you some information about how to find this job? 5. If you look for work you need to provide referees so that the employer can obtain some information about you. Whom did you cite the last time you looked for a job or an apartment? Whom would you ask if you did have to? 6. Have you ever borrowed money? Whom did you turn to? If you needed money, whom would you ask? 7. Is there anybody who asks you for money when they need it? Did you assist anybody financially during the last year? 8. With whom have you shared social activities lately? With whom do you spend your leisure time? This means spending the weekends together, going for a drink or to the cinema. 9. Are there people outside of Switzerland who are important for you and whom you have not yet mentioned? 10. We will now have a look at the persons mentioned: Is there someone else who is important for you who is not on this list? If there is, I would like to know his name and add it to list. 4) It should be noted that Switzerland accords citizenship on a ius sanguinis basis, with the result that many of the people counted as foreign nationals were born in Switzerland, sometimes even of parents also born in Switzerland. Data are derived from the Ministry of Population Service or from the most recent census (2000). 5.) The statistical analysis shows that the variables measure indeed three different aspects of network diversity confirming hereby the theoretically motivated choice of the variables. Spearman’s rank correlation coefficient makes evident that the correlation between the three variables is small or medium. 6.) I did run some regression models to examine the factors that might explain network diversity and range. However, those results remained very unsatisfactory and I decided to use those complex samples procedures that are more meaningful.

BIBLIOGRAPHY Alba, Richard (2005). "Bright vs. blurred boundaries: Second-generation assimilation and exclusion in France, Germany, and the United States." Ethnic and Racial Studies, 28(1): 20-49. Alba, Richard D. und Nee Viktor (1997). "Rethinking Assimiliation Theory for a New Era of Immigration." International Migration Review, 31(4): 826-874. Barth, Fredrik (1969). "Introduction", in Barth, Frederik (ed.) Ethnic Groups and Boundaries: The Social Organization of Culture Difference. London: Allen & Unwin, p. 9-38. Beck, Ulrich (2002). Macht und Gegenmacht im globalen Zeitalter. Neue weltpolitische Ökonomie. Frankfurt am Main: Suhrkamp Verlag. Berset, Alain and Olivier Crevoisier (2006). "Circulation of Competencies and Dynamics of Regional Production Systems." International Journal on Multicultural Societies (IJMS), 8(1): 61-83. Bourdieu, Pierre (1980). "Le capital social. Notes provisoires." Actes de la recherche en sciences sociales, 31: 2-3.

18

(1982). Die feinen Unterschiede. Kritik der gesellschaftlichen Urteilskraft. Frankfurt am Main: Suhrkamp. Brubaker, Rogers (2004). "Ethnicity without Groups", in Brubaker, Rogers (ed.) Ethnicity without Groups. Cambridge: Harvard University Press, p. 7-27. Burt, Ronald S. (1982). Towards a Structural Theory of Action. Network Models of Social structure, Perception and Action. New York: Academic Press. Burt, Ronald S (1983). "Range", in Burt, Ronald S and Michael J Minor (eds.) Applied Network Analysis: A Methodological Introduction. Beverly Hills: Sage, p. 176-194. Charmaz, Kathy (2001). "Qualitative Interviewing and Grounded Theory Analysis", in Gubrium, Jaber F and James A Holstein (eds.) Handbook of Interview Research. Context and Methods. Thousand Oaks: Sage Publications, p. 675-694. Coleman, James S (1988). "Social capital in the Creation of Human Capital." The American Journal of Sociology, 94: 95-120. Dahinden, Janine (2005). "Contesting transnationalism? Lessons from the study of Albanian migration networks from former Yugoslavia." Global Networks. A journal of transnational affairs, 5(2): 191-208. Dahinden, Janine (2008). "Deconstructing Mythological Foundations of Ethnic Identities and Ethnic Group Formation: Albanian-Speaking and New Armenian Immigrants in Switzerland." Journal of Ethnic and Migration Studies, 34(1): 55 - 76. Dahinden, Janine (2009). "Are we all transnationals now? Network transnationalism and transnational subjectivity: the differing impacts of globalization on the inhabitants of a small Swiss city." Ethnic and Racial Studies, 32 (8): 1365-1386. Degenne, Alain und Michel Forsé (1994). Les réseaux sociaux: Une analyse structurale en sociologie. Paris: Arman Colin. Elias, Norbert and John L Scotson (1965). The Established and the Outsiders. A Sociological Enquiry into Community Problems. London: Frank Cass & Co. Flap, Henk and Beate Völker, eds. (2004). Creation and Returns of Social Capital. London & New York: Routledge. Flap, Henk et al. (2005). Measurement instruments for social capital of individuals. Questionnaire items as used in the 1999/2000 nationa wide study "Social relations and networks in the neighborhood and at the workplace: The Social survey of the Networks of the Dutch (SSND) – a joint project of the universities of Utrecht, Groningen and Amsterdam. Glaser, Barney and Anselm Strauss (1967). The Discovery of the Grounded Theory: Strategies for Qualitative Research. Chicago: Aldine Publishing Company. Glick Schiller, Nina and Ayse Caglar (2009). "Towards a Comparative Theory of Locality in Migration Studies: Migrant Incorporation and City Scale." Journal of Ethnic and Migration Studies, 35(2): 177-202. Gordon, Milton (1964). Assimilation in American life. New York: Oxford University Press. Granovetter, Mark S. (1973). "The Strength of Weak Ties." American Journal of Sociology, 78(6): 1360-1380. Hanley, Lisa M, Blair A Ruble and Allison M Garland, eds. (2008). Immigration and Integration in Urban Communities. Renegotiating the City. Baltimore: The Johns Hopkins University Press. Jenkins, Richard (1997). Rethinking Ethnicity: Arguments and Explorations. London: Sage. Lamont, Michèle and Virag Molnar (2002). "The Study of Boundaries in the Social Sciences." Annual Review of Sociology, 28: 167-195.

19

Lazarsfeld, Paul und Robert K Merton (1954). "Friendship as a Social Process: A Substantive and Methodological Analysis", in Berger, Morroe (Hg.), Freedom and Control in Modern Society. New York: Van Nostrand, S. 18-66. Lin, Nan (1999). "Building a Network Theory of Social Capital." Connections, 22(1): 28-51. Marsden, Peter V. (1990). "Network Data and Measurement." Annual Review of Sociology, 16: 435-463. McPherson, Miller, Lynn Smith-Lovin and James M Cook (2001). "Birds of a Feather: Homophily in Social Networks." Annual Review of Sociology, 27: 415-444. Meyer, Jean-Baptiste (2001). "Network Approach versus Brain Drain: Lessons from the Diaspora." International Migration, 39(5): 91-108. Mitchell, J. Clyde (1969). Social Networks in Urban Situations. Analysis of Personal Relationships in Central African Towns. Manchester: University of Manchester at the University Press. Moore, Gwen (1990). "Structural Determinants of Men's and Women's Personal Networks." American Sociological Review, 55(5): 726-735. Piguet, Etienne (2005). L'immigration en Suisse depuis 1948. Une analyse des flux migratoires vers la Suisse. Zürich: Seismo. Portes, Alejandro (1998). "Social Capital: Its Origins and Applications in Modern Sociology." Annual Review of Sociology, 24(1): 1-24. Portes, Alejandro und Rubén G Rumbaut (1996). Immigrant America : A Portrait. Berkeley [etc.]: University of California Press. Schweizer, Thomas (1996). Muster sozialer Ordnung. Netzwerkanalyse als Fundament der Sozialethnologie. Berlin: Dietrich Reimer Verlag. Scott, John (1991). Social Network Analysis. A Handbook. London: Sage. Van der Gaag, Martin und Tom Snijders (2004). "Proposal for the Measurement of Individual Social Capital", in Flap, Henk und Beate Völker (Hg.), Creation and Returns of Social Capital. A New Research Program. London and New York: Routledge, S. 199-218. Vertovec, Steven (2007). "Super-diversity and its implications." Ethnic and Racial Studies, 30(6): 1024-1054. Weber, Max (1991 [1904]). "Die "Objektivität" sozialwissenschaftlicher und sozialpolitischer Erkenntnis", in Sukale, Michael (ed.) Max Weber. Schriften zur Wissenschaftslehre. Stuttgart: Philipp Reclam, p. 21-101. (1996 [1922]). "Ethnic Groups", in Sollors, Werner (ed.) Theories of Ethnicity. A Classical Reader. [from Chapter V in Guenther Roth and Claus Wittich (eds), Economy and Society (Berkeley: University of California Press 1978), pp. 385-98, originally published in 1922. New York: New York University Press, p. 52-66. Wimmer, Andreas (2004). "Does Ethnicity matter? Everyday group formation in three Swiss immigrant neighbourhoods." Ethnic and Racial Studies, 27(1): 1-36. (2009). "How to study ethnicity in immigrant societies. Herder's heritage and the boundary-making approach." Sociological Theory, 27. Wimmer, Andreas and Nina Glick Schiller (2002). "Methodological Nationalism and Beyond: Nation-State Building, Migration and the Social Sciences." Global Networks, 2(4): 301-334. Zhou, Min (1997). "Segmented Assimilation: Issues, Controversies, and recent Research on the new Second Generation." International Migration Review, 31(4): 975-1008.

20